Introduction

The abilities of Shigella to enter colonic epithelial cells and subsequently to spread within and between the cells are a prerequisite for causing dysentery. The capacity of the bacteria to spread in the cytoplasm and then to move into adjacent epithelial cells is mediated by the bacterial functions encoded by the virG (icsA) gene (Makino et al., 1986; Bernardini et al., 1989; Lett et al., 1989). After invasion of epithelial cells, Shigella become surrounded by filamentous actin (F-actin). This F-actin 'clot' then rearranges into a tail (also known as a 'comet') which remains stationary in the cytoplasm and is left behind by bacteria moving ahead (Bernardini et al., 1989; Prévost et al., 1992; Goldberg et al., 1993). When the spreading bacterium contacts the inner surface of the host cell plasma membrane, a long membrane protrusion (filopodium) develops with the F-actin tail behind it that is endocytosed by the adjacent cells, resulting in the bacterium being surrounded by a double membranous vacuole (Kadurugamuwa et al., 1991; Allaoui et al., 1992; Goldberg et al., 1993; Suzuki et al., 1994). After disruption of the vacuole, bacteria are released into the new host cell cytoplasm and multiply again. Thus, this actin-based bacterial motility is critical for allowing invading Shigella to spread horizontally among colonic epithelial cells.

VirG is a surface-exposed outer membrane protein, and the sequence of the virG gene revealed that VirG is a 1102 amino acid protein (Lett et al., 1989). VirG protein consists of three distinctive domains, the N-terminal signal sequence (amino acids 1–52), the 706 amino acid -domain (amino acids 53–758) and the 343 amino acid C-terminal -core (amino acids 759–1102) (Goldberg et al., 1993; Suzuki et al., 1995). The -domain is presented on the surface of Shigella and contains repetitive sequences rich in glycine residues, while the -core is embedded in the outer membrane (Suzuki et al., 1995). The assembly of F-actin near the surface of intracellular Shigella is absolutely dependent on the surface presentation of VirG -domain, and the formation of an actin tail depends on the asymmetric distribution of VirG (Goldberg et al., 1993; Rajakumar et al., 1994; Sandlin et al., 1995; Suzuki et al., 1996). Recently, to demonstrate that VirG is the sole bacterial factor needed for movement in infected mammalian cells, experiments on actin-based movement were carried out using Xenopus laevis egg extracts (Goldberg and Theriot, 1995; Kocks et al., 1995). In these studies, heterologous VirG expression in Escherichia coli K-12 was achieved by exploiting an OmpT^- (OmpT is a trypsin-like outer membrane protease) VirG-expressing strain, in which the cleavage of VirG and, thus, the release of the -domain into the bacterial environment, was prevented (Nakata et al., 1993). This strain displayed an asymmetric distribution of VirG on the bacterial body similar to Shigella, and was capable of intra- and intercellular movement (Kocks et al., 1995).

The movement of S.flexneri (and Listeria monocytogenes) within the host cells resembles the formation of filopodia of locomoting cells, since those are mediated by actin polymerization and cause membrane protrusions (Bernardini et al., 1989; Prévost et al., 1992; Nobes and Hall, 1995). However, no known protein has yet been reported to be involved specifically in both phenomena. Several host proteins such as vinculin (Kadurugamuwa et al., 1991; Suzuki et al., 1996; Laine et al., 1997), plastin (fimbrin) (Prévost et al., 1992), filamin (Prévost et al., 1992), -actinin (Zeile et al., 1996) and vasodilator-stimulated phosphoprotein (VASP) (Chakraborty et al., 1995) have been identified as being associated with the F-actin tail generated by intracellular S.flexneri. Amongst these proteins, only vinculin can interact directly with VirG (Suzuki et al., 1996; Laine et al., 1997), in which the 95 kDa vinculin head domain interacting with the VirG -domain is involved in promoting the formation of actin tail from intracellular Shigella. However, vinculin is not only associated with Shigella actin tail but is also distributed widely in focal adhesions, filopodia and lamellipodia. Recently, neural Wiskott–Aldrich syndrome protein (N-WASP) was identified as the Ash/Grb2-binding protein in brain, but its expression was also noted in other tissues including the colon (Miki et al., 1996). The sequence of N-WASP is homologous to WASP (Derry et al., 1994), and they both possess various functional motifs, such as a pleckstrin-homology (PH) domain, an IQ motif, a GBD/CRIB motif, a proline-rich region (P-region), a verprolin-homology region (V-region) and an ADF/cofilin-homology region (C-region) (Miki et al., 1996). Importantly, when N-WASP or WASP is overexpressed in mammalian cells, only N-WASP induces long filopodia (Miki et al., 1998). These data led us to investigate whether N-WASP is involved directly in actin comet formation by intracellular Shigella.

Results

Intracellular S.flexneri can recruit N-WASP at one pole of the bacterium

HeLa, Cos-7, PtK2 or Caco-2 cells and Xenopus egg extracts, which all have been used for studies of bacterial motility, were investigated for their expression of N-WASP. Although expression of N-WASP in each cell lysate, as assayed by an immunoblot with an anti-N-WASP antibody, varied slightly, they expressed significant amounts of the protein (data not shown). HeLa cells infected with wild-type S.flexneri 2a strain YSH6000 were examined further for the cellular distribution of N-WASP by triple immunostainings with rhodamine–phalloidin, Cy2-labeled rabbit anti-N-WASP antibody and Cy5-labeled rat anti-VirG antibody. After 60 min of infection, N-WASP was highly concentrated at one pole of the bacterium, from which F-actin was also assembled as a comet tail (Figure 1A, B and D). Under the same conditions, N-WASP did not accumulate at all on the surface of M94, a virG-deficient derivative of YSH6000 unable to form an actin comet tail (data not shown). Indeed, the N-WASP accumulated at one pole of intracellular YSH6000 was always confined to the area expressing VirG on the bacteria (Figure 1C). Similar results were obtained with human Caco-2 cells (data not shown). To assess further the role of N-WASP in the Shigella-induced actin tail, HeLa cells were infected with YSH6000 for 20, 40 and 60 min, and then examined for changes in the distributions of N-WASP, VirG and F-actin on the bacteria by confocal microscopy. After 20 min of infection, all intracellular YSH6000 accumulated VirG at one pole of bacterium and approximately half of the bacteria had detectable F-actin and N-WASP accumulation. At 60 min after infection, 90% of bacteria possessed an assembled F-actin. At this time point, the N-WASP associated with Shigella was still confined to the front of the tail, suggesting that N-WASP plays a role in assembly of F-actin in infected cells.

N-WASP cannot be recruited by intracellular Listeria

Like Shigella, L.monocytogenes, a Gram-positive bacterium, moves in mammalian cells by generating an actin comet (for reviews, see Lasa and Cossart, 1996; Machesky, 1997), an ability which requires the expression of ActA protein on the bacterial surface (Domann et al., 1992; Kocks et al., 1992). Hence, we infected HeLa cells with L.monocytogenes 1/2a EGD and examined its ability to recruit N-WASP on the bacterial surface by double immunostaining with Cy2-labeled anti-N-WASP antibody and rhodamine–phalloidin. As shown in Figure 1E–G, at 120 min after infection, none of the bacteria in the HeLa cells had recruited localized N-WASP in the region of the actin comet tail, suggesting that accumulation of N-WASP at one pole of the intracellular bacterium is a Shigella-specific event.

Shigella VirG binds to the verprolin-homology region of N-WASP

Since N-WASP accumulated at the site of VirG expression on intracellular Shigella, the ability of N-WASP to interact with VirG was first investigated by co-precipitation experiments. HeLa cell lysates were incubated with GST–1 and GST–Ash/Grb2 fusion proteins or GST alone immobilized to glutathione–Sepharose. GST–1 contained residues Thr53–Arg779 of the VirG peptide, an essential portion for eliciting actin comet tail formation (Suzuki et al., 1996), while GST–Ash/Grb2 (Lowenstein et al., 1992; Matuoka et al., 1992) was exploited originally for the isolation of N-WASP (Miki et al., 1996). Proteins co-precipitating with the GST beads were analyzed by immunoblots with anti-N-WASP antibody. As shown in Figure 2B, N-WASP co-precipitated with GST–1 as well as with GST–Ash/Grb2, but not with GST alone, suggesting that N-WASP was capable of interacting with VirG in vitro. To assess which domains in N-WASP are responsible for interacting with VirG, Cos-7 transfectants (Figure 2A), that overexpressed either the full-length N-WASP (Met1–Asp505), VCA protein (lacking the verprolin- and cofilin-homology regions) or V protein (lacking the verprolin-homology region), were lysed and incubated with GST–1. The proteins precipitated by GST–1 beads were then analyzed by immunoblotting with anti-N-WASP antibody. The full-length N-WASP, but not the other products, VCA and V, were precipitated with GST–1 (Figure 2C), indicating that the verprolin-homology region of N-WASP could be involved in the interaction with VirG.

To investigate further whether this verprolin-homology region can interact directly with VirG, we constructed four GST fusion proteins containing various portions of the VCA region of N-WASP (Figure 2A) and examined their ability to interact with the VirG -domain. The resulting GST fusion proteins were blotted onto a membrane filter, overlaid with VirG-1 (see Materials and methods), and its binding to the VCA segments was examined by immunostaining with rabbit VRG-N2 antibody (a VirG-specific antibody). The results showed that GST–VCA, GST–VC and GST–V, but not GST–CA or GST alone, were bound by VirG-1 (Figure 2D). We thus concluded that the 58 amino acids of the verprolin-homology region of N-WASP take part in the interaction with VirG.

The verprolin-homology region of N-WASP interacts with VirG in vivo

To investigate whether the verprolin-homology region of N-WASP is able to interact with VirG in vivo, PtK2 cells infected with S.flexneri M94 (a virG mutant of YSH6000) carrying pD10-1, a plasmid expressing VirG at a high level (Suzuki et al., 1996), were microinjected with purified GST–V. Fifteen minutes after microinjection, the PtK2 cells were immunostained with Cy2-labeled goat anti-GST antibody, Cy5-labeled rabbit VRG-N2 antibody and rhodamine–phalloidin. As shown in Figure 3, the injected GST–V was accumulated around the VirG-expressing area on Shigella (Figure 3A and B), where the actin tail was generated (Figure 3C). When GST–V was injected into Ptk2 cells infected with L.monocytogenes, it did not accumulate at all on the intracellular bacteria (data not shown). These data further confirmed that the verprolin-homology region of N-WASP is involved in specific interaction with VirG expressed on intracellular Shigella.

Identification of the VirG domain interacting with N-WASP

Our recent studies have indicated that the VirG -domain (the surface-exposed VirG sequence) consists of two distinct regions, an N-terminal portion corresponding to amino acids 53–508, and the remaining C-terminal portion corresponding to amino acids 509–758 (Suzuki et al., 1995, 1996). The N-terminal region contains eight glycine-rich repeats, and is essential for eliciting an actin comet tail in intracellular Shigella, while the following C-terminal region is required for the unipolar deposition of VirG on the bacterium (Suzuki et al., 1996). In order to identify which VirG portion takes part in the interaction with N-WASP, we constructed various GST–-domain derivatives, named GST–1 to GST–8 (see Figure 4A), and tested each of them for their ability to interact with N-WASP by incubating them in HeLa cell lysates followed by precipitation with the beads to which the GST derivatives were attached. Analysis of the precipitated proteins by immunoblotting with anti-N-WASP antibody indicated that GST–1, GST–3 and GST–7 constructs containing all of the eight glycine-rich repeats could be bound by N-WASP (Figure 4B). To confirm that these three VirG -domain portions were capable of binding N–WASP in infected cells, Shigella strain M94 carrying pD10-1 (the whole VirG -domain), pD10-virG3 (VirG3, lacking amino acids 509–729 of the VirG -domain) or pD10-virG4 (VirG4, lacking amino acids 104–506 of the VirG -domain) (Suzuki et al., 1996) was examined for its ability to recruit an N-WASP clot as well as recruit an F-actin assembly on the intracellular bacterium. The results showed that bacteria expressing wild-type VirG and VirG3 but not VirG4 were capable of eliciting accumulation of N-WASP and F-actin assembly in the infected cells as examined by immunostaining (data not shown). These results indicated that the minimum VirG domain required for the interaction with N-WASP lies within the Arg103–Ala433 region in the VirG polypeptide.

A dominant-negative N-WASP affects actin assembly on intracellular Shigella

It has been indicated that activated N-WASP promotes the depolymerization activity of actin filaments through its ADF/cofilin-homology region, and that overexpression of cof, a mutant lacking four amino acids highly conserved in cofilin-family members, can inhibit filopodium formation, thus indicating that the cof mutant functions in a dominant-negative manner (Miki et al., 1998). Thus, we exploited the cof mutant to check the observed interaction of N-WASP with VirG involved in the assembly of F-actin on Shigella in infected mammalian cells. As shown in Figure 5A, the cof mutant was capable of interacting with VirG in vitro. The effect of overexpression of the cof mutant on the formation of actin comet tail by Shigella infecting Cos-7 cells was examined by counting the number of actin assemblies associated with the intracellular bacteria in Cos-7 cells overexpressing cof or the full-length N-WASP. As shown in Figure 5B and C, after 60 min of infection, the intracellular bacterium had an actin tail in Cos-7 cells overexpressing the full-length N-WASP, whereas it was almost abolished in Cos–7 cells overexpressing the cof mutant. Approximately 90% (92.8 7.2%, n = 85) of intracellular bacteria possessed an actin clot in Cos-7 cells overexpressing the full-length N-WASP, whereas only 16% (16.4 4.9%, n = 138) had an actin assembly in Cos-7 cells overexpressing the cof mutant, even though cof accumulation at one pole of the bacteria was observed (Figure 5C, panel b). The overexpression of cof or the full-length N-WASP had no effect on actin comet formation from intracellular L.monocytogenes (Figure 5B). These results show that the overexpression of a dominant-negative N–WASP mutant greatly inhibits formation of actin tail by intracellular S.flexneri.

Depletion of N-WASP from Xenopus egg extracts affects the ability of VirG-expressed bacteria to recruit actin assembly

To investigate further the functional requirement for N–WASP for the assembly of an actin tail by intracellular Shigella, we depleted N-WASP from Xenopus egg extracts by immunoprecipitation with anti-N-WASP antibody immobilized on protein A beads. The immunodepleted extracts were then tested for their ability to support the possession of an actin clot or tail from E.coli MC1061 ompT::Km carrying pD10-1, according to the methods described by Goldberg and Theriot (1995). Upon depletion of N-WASP, generation of an actin tail was shut off, and only 17.6 5.0% (n = 229) of the bacteria associated with a weak actin clot around the bacterial body (Figure 6B and C, panel b), whereas 80% of bacteria (82.9 2.3%, n = 198) possessed an actin clot or tail in a mock-depleted extract (Figure 6B and C, panel a), this level being similar to that in the untreated extract (83.7 4.8%, n = 135). When a similar amount of N–WASP, prepared by a baculovirus system, to that in the original extracts (45 nM) was added to the N-WASP-depleted extracts, although the formation of a long actin tail from the bacteria was less efficient than in the original extract (Figure 6C, panels a and c), the bacterial population associated with an actin clot was greatly increased (88.0 1.1%, n = 233) (Figure 6B and C, panel c). These data demonstrate that N-WASP is functionally involved in the promotion of actin assembly by Shigella.

Discussion

Recent studies have shown that some bacteria such as Shigella, Listeria and Rickettsia or viruses such as vaccinia virus have gained an ability to control dynamic remodeling of the actin cytoskeleton in order to achieve actin-based motility in mammalian cells in their efforts to spread within and between the cells and evade the host defense system (Lasa and Cossart, 1996; Higley and Way, 1997; Machesky, 1997). Although the precise mechanisms underlying actin-based motility still remain to be elucidated, the systems exploited by such pathogens have provided a simplified model that mimics the actin filament dynamics involved in lamellipodial or filopodial protrusions. To gain further insights into Shigella actin-based motility, we investigated the possibility of N-WASP involvement in actin tail formation by intracellular Shigella, since N–WASP has been shown to bind actin directly, depolymerize actin filaments and regulate cortical cytoskeletal rearrangement, including formation of filopodia (Miki et al., 1996, 1998). Furthermore, although N-WASP is expressed predominantly in brain, it is also distributed in other tissues including the colon (Miki et al., 1996). In this study, we found that N-WASP can interact with VirG under in vitro and in vivo conditions and plays a critical role in eliciting Shigella actin tail formation in infected cells. This conclusion was deduced from the following results: (i) N-WASP accumulation at one pole of intracellular Shigella is always confined to the VirG deposition area on the bacterial surface; (ii) N-WASP is able to interact with a VirG domain (-domain), which we previously showed to be required for assembly of F-actin on intracellular Shigella (Suzuki et al., 1996); (iii) overexpression of a dominant-negative N-WASP mutant in Cos-7 cells infected with Shigella resulted in a decrease in actin tail formation; and (iv) actin tail formation by Shigella supported by Xenopus egg extracts was shut off upon depletion of N-WASP.

N-WASP contains at least six distinct domains: a PH domain; an IQ motif; a GBD/CRIB motif; a proline-rich region; a verprolin-homology region; and an ADF/cofilin-homology region (Miki et al., 1996). The proline-rich region (P-region), verprolin-homology region (V-region) and ADF/cofilin-homology region (C-region) are likely to participate directly in generation of the actin tail from the surface of intracellular Shigella. The V-region was shown to be able to bind directly to VirG under in vitro and in vivo conditions. The amino acid sequence of the P-region, which contains four Gly(Pro)5 sequences, was equivalent to the putative profilin-binding motifs distributed among the VASP (Haffner et al., 1995), WASP (Derry et al., 1994) and Mena (Gertler et al., 1996) proteins, although the number of motifs varied. Since it has been indicated that those proline-rich repeats interact with profilin (Reinhard et al., 1995; Gertler et al., 1996; Higley and Way, 1997), and the P-region of N-WASP can interact with profilin in vitro (Miki et al., 1998), we presume that N-WASP-bound profilin could mediate local actin polymerization on the surface of intracellular Shigella (Figure 7). Recent studies have indicated that activation of N-WASP by binding of Cdc42 to the GBD/CRIB motif results in exposure of the VCA actin-depolymerizing region, which is thought to be masked in unstimulated full-length N-WASP through the intramolecular interaction between the acidic residues at the C-terminal end and the basic region near the GBD/CRIB motif region (Miki et al., 1996, 1998). In this context, the VCA region of N-WASP may be important for the formation of actin tail from intracellular Shigella, in that the stable capped actin filaments may be partially depolymerized by N-WASP and the uncapped barbed ends become exposed to serve as sites for actin polymerization (Zigmond, 1996; Welch et al., 1997). At the uncapped barbed ends, actin polymerization would occur and it could be promoted by profilin (Pantaloni and Carlier, 1993).

It has been postulated that host cytoskeletal proteins responsible for promoting actin filament nucleation and elongation from intracellular Shigella and Listeria should be localized only at the front of the tail, and perhaps associated with the surface of the bacterium (Sanger et al., 1992; Theriot et al., 1992; Goldberg and Theriot, 1995). Hence, we investigated the distribution of N-WASP together with that of VASP in Shigella actin tail in infected cells and compared it with that of intracellular Listeria. Although the N-WASP clot appearing on intracellular Shigella was always immediately adjacent to the back half of the bacterium, it was never detected on intracellular Listeria. However, VASP has been reported to be accumulated around the area expressing ActA on intracellular L.monocytogenes, and the VASP protein was shown to bind directly to the proline-rich repeats domain in ActA (Niebuhr et al., 1997). In contrast to Listeria, VASP was shown to be unable to bind to VirG in vitro, perhaps due to the absence of proline-rich repeats in the VirG polypeptide (Chakraborty et al., 1995). Examination of the distribution of VASP in HeLa cells infected by S.flexneri or L.monocytogenes using immunostaining confirmed that the distribution of VASP in the actin tail generated by Shigella and Listeria was different (T.Suzuki, unpublished data). In the experiment, we noted that the distribution of VASP, but not of N-WASP, was similar to the distribution of vinculin in the Shigella actin tail (Suzuki et al., 1996; Laine et al., 1997). Since it has been shown that VASP is bound by the proline-rich domain in vinculin (Brindle et al., 1996; Reinhard et al., 1996), it is thus possible that vinculin and VASP could be co-localized in the actin tail in Shigella-infected cells. Based on our results together with other studies (Chakraborty et al., 1995; Zeile et al., 1996), we assume that the role of VASP in the assembly of actin in Listeria is different from that in Shigella.

In this study, we have identified the VirG region involved in interacting with N-WASP, which corresponds to Arg103–Ala433 in the VirG polypeptide. This region contains eight glycine-rich repeats (G-region) and is essential for recruiting F-actin assembly on intracellular Shigella (Suzuki et al., 1996). In our previous study, we found that the G-region of VirG had a capacity to interact with the vinculin head in vitro. Indeed, we observed that the ability of the vinculin head to bind to to GST–1, which contains the G-region of VirG, was reduced upon addition of the V-region of N-WASP in a dose-dependent manner in vitro, suggesting that the domains in VirG binding to N-WASP and vinculin could overlap at least in part (T.Suzuki, unpublished data). The vinculin head has been shown to bind talin, while the vinculin tail portion can bind and cross-link F-actin (Johnson and Craig, 1994, 1995). Microinjection studies have indicated that injection into PtK2 cells infected with S.flexneri of one of the proline-rich sequences of ActA (FEFPPPPTDE) or that of VASP (GPPPPPGPPPPPGPPPPP) results in decreased Shigella movement that is restored upon co-injection of profilin (Zeile et al., 1996). Therefore, in this way, vinculin may serve as an adaptor protein for VirG on intracellular bacteria, in which the vinculin may recruit VASP-bound profilin or another host protein that facilitates elongation from the barbed end of nascent actin filaments at the bacterial surface (Figure 7). Alternatively, the VirG-bound vinculin might serve as an actin filament recruiter (Johnson and Craig, 1994; Jockusch and Rüdiger, 1996) so that actin filament elongation on the Shigella surface with the aid of other VirG-bound host proteins such as N-WASP would be feasible (Figure 7). Since the affinity of GST–1 for the vinculin head portion, as measured by its K_d value (0.3 M) (T.Suzuki, unpublished data), is similar to the affinity of the vinculin tail for F-actin (0.6–0.8 M) (Johnson and Craig, 1995), it is possible that the vinculin, once having interacted with VirG, could in turn be diffused into the elongated actin tail, leaving the N-WASP bound by VirG on the Shigella surface. Although we have discussed an important role for vinculin in Shigella actin-based motility, there have been two controversial reports on the requirement for vinculin. Goldberg (1997) examined the functional role of vinculin in Shigella actin-based motility in a vinculin-deficient F9 cell line variant 5.51, and observed that Shigella are able to form actin tails and are motile as in the F9 cells, suggesting that vinculin is not the essential factor. In contrast, Laine et al. (1997) found that vinculin proteolysis occurred in Shigella-infected cells and that the cleaved head portion of vinculin is a rate-limiting factor in Shigella motility. In the latter study, they proposed that the cleaved vinculin head bound by VirG can interact with VASP, through which profilin could be recruited, thus promoting assembly of actin on the Shigella surface. In this regard, it should be noted that cleavage of vinculin in HeLa cells infected by S.flexneri 2a YSH6000 or M94 (a virG mutant of YSH6000) strains was not detected after 60 min of infection in our study (T.Suzuki, unpublished data). At present, therefore, we are not sure how these discrepant observations were made in each study, whether they merely reflect the different cell types or the absence of appropriate control. In any case, we must await further studies to elucidate the exact role of VirG-bound vinculin and N-WASP in generating a Shigella actin tail.

In this study, we performed the functional assay for the requirement for N-WASP for the assembly of an actin clot and tail by intracellular Shigella by using Xenopus egg extracts, and revealed that depletion of N-WASP from the Xenopus egg extracts affected the ability of VirG-expressed bacteria to recruit actin assembly. Although the purified N-WASP was sufficient to support the assembly of dense actin clouds and sometimes a short actin tail (insets in Figure 6C, panel c), it was still less efficient than in the original extract for the formation of a long actin tail by the bacteria. This may result from the elimination of other host factor(s) required for facilitating the elongation of actin tail which could have been co-precipitated together with N-WASP, or it may result from the use of heterogenous N-WASP which was added to the depleted extract (see Materials and methods).

In conclusion, our results show that N-WASP is required to trigger both actin assembly and movement of Shigella in mammalian cells, although this was not the case in intracellular Listeria. Although the precise role of N–WASP in assembly of actin filaments on the Shigella surface still awaits further study, the results presented in this study will provide some insight into understanding the systems underlying actin-based motility of other pathogens as well as the formation of filopodia from locomoting mammalian cells.

Materials and methods

Bacterial strains, cell culture and media

The S.flexneri 2a YSH6000 and L.monocytogenes serotype 1/2a EGD strains have been described previously (Sasakawa et al., 1986; Domann et al., 1992). M94 (S.flexneri virG::Tn5) carrying pD10-1, pD10-1virG3 or pD10-1virG4 were prepared as described previously (Suzuki et al., 1996). All strains were grown routinely in brain–heart infusion (BHI) broth (Difco) at 37°C. HeLa cells were maintained in minimal essential medium (MEM) (Nissui, Tokyo, Japan) with 10% fetal calf serum (FCS) (Nichirei, Tokyo, Japan), Caco-2 cells were grown in MEM supplemented with 10% FCS and 0.1 mM non-essential amino acids (Gibco), and Cos–7 cells and PtK2 cells (ATCC CCL 56) were grown in Dulbecco's modified Eagle's medium (DMEM) (Sigma) containing 10% FCS.

Antibodies

The anti-N-WASP-specific rabbit antibody has been described previously (Miki et al., 1996). The rabbit VRG-N2 and anti-GST antibodies were prepared as described by Suzuki et al. (1996). The rat VRG-N2 was obtained by immunization of rats with the same synthesized peptides as for rabbit antibody.

Infection of cultured cells and immunofluorescence microscopy

All cells were grown on glass coverslips to 60% confluency in the absence of antibiotics. Cells were infected with bacteria at a multiplicity of infection of 100 per cell. In the case of S.flexneri, the plates were centrifuged at 700 g for 10 min after adding bacteria. After 20–60 min, the plates were washed extensively with phosphate-buffered saline (PBS) and fresh medium supplemented with gentamycin (100 g/ml) and kanamycin (60 g/ml) added. The infected cells were incubated for an additional 1 h (S.flexneri) or 2 h (L.monocytogenes) at 37°C before fixation in 4% paraformaldehyde in PBS for 20 min. The coverslips were incubated in 50 mM NH₄Cl in PBS for 10 min and permeabilization carried out in 0.2% Triton X-100 in PBS for 20 min. After blocking for 30 min in 2% bovine serum albumin (BSA) in Tris-buffered saline (TBS), the coverslips were incubated with rat VRG-N2 or anti-N-WASP antibody in TBS. CyDye-conjugated secondary antibodies (Amersham) were used to visualize VirG and N-WASP, and rhodamine–phalloidin (Molecular probes) was used to visualize F-actin. The coverslips were mounted in Vectashield (Vector) and observed with a confocal laser-scanning microscope (MRC-1024, Bio-Rad). Accumulation of N-WASP or F-actin around the intracellular bacteria was defined as areas where the intensity was higher than an arbitrary threshold of 80 estimated in non-infected cells by using image processing software (LaserSharp version 2.0, Bio-Rad).

GST fusion proteins

GST–Ash/Grb2 and GST–VirG -domain fusion protein series GST–1 to 6 were constructed as described previously (Miki et al., 1994; Suzuki et al., 1996). GST fusion proteins of N-WASP were constructed as previously described (Miki et al., 1996). For preparation of GST–7 and 8, the DNA fragments encoding the amino acids shown in Figure 4A were generated by PCR and cloned into pGEX-2T (Pharmacia). These GST fusion proteins were purified using protocols supplied by the manufacturer. For use in the gel blot overlay assay, the GST fusion proteins were cleaved with 0.4 mg/ml human thrombin (Sigma) in thrombin buffer [100 mM NaCl, 50 mM Tris–HCl, pH 7.5, 2.5 mM CaCl₂, 5 mM MgCl₂, 1 mM dithiothreitol (DTT)] at 37°C for 1 h. The thrombin was removed by adding p-aminobenzamidine–agarose beads (Sigma) for 30 min at 4°C. Purified proteins were dialyzed against TBS and concentrated by ultrafiltration using Centricon-10 (Amicon).

In vitro binding assay using GST fusion proteins

HeLa and Cos-7 cells grown in 100 mm dishes were lysed in ice-cold RIPA buffer [25 mM Tris–HCl, pH 7.5, 150 mM NaCl, 1% NP-40, 0.5% deoxycholate, 0.1% SDS, 1 mM Na₃VO₄, 1 mM phenylmethylsulfonyl fluoride (PMSF), 10 g/ml leupeptin]. Recombinant GST fusion proteins immobilized on glutathione–Sepharose were incubated with 1 ml of lysate overnight at 4°C. Samples were washed three times in RIPA buffer without deoxycholate and SDS and subjected to immunoblotting. For gel blot overlay assays, proteins to be analyzed were electrophoresed on an SDS–polyacrylamide gel followed by blotting onto a nitrocellulose membrane. After incubating in blocking solution (2% BSA, 0.05% sodium azide in TBS), purified recombinant VirG-1 (10 g/ml) in blocking solution was incubated with the filter. After washing with TBS–0.05% Tween-20, bound protein was detected with rabbit VRG-N2 antibody.

Transient expression of N-WASP and its mutants in Cos-7 cells

Full-length and VCA constructs were prepared as described previously (Miki et al., 1996). For preparation of V and cof constructs, the DNA fragments encoding the amino acids shown in Figure 2A were generated by PCR and cloned into pcDL-SRII mammalian expression vector. A total of 0.510⁷ cells were mixed with each purified plasmid (20 g) and transfected by electroporation. The cells were re-plated and cultured for 48 h before bacterial infection. Transfected cells were defined as cells where the total cellular intensity of N-WASP was >10-fold higher than that of non-transfected cells.

Microinjection

Purified GST–V fusion protein was microinjected into the cytoplasm of PtK2 cells infected with bacteria. The needle concentration of proteins was 100 M, and 10% of the total cell volume was injected. Cells were then returned to the incubator for 15 min before fixation.

Actin tail assay in Xenopus egg extracts

Meiotically arrested cytoplasmic extracts of X.laevis were prepared as described previously (Theriot et al., 1994; Rosenblatt et al., 1997). Pre-immune rabbit IgG or anti-N-WASP antibody was bound to 30 l of protein A-conjugated Sepharose 4B (Sigma) in TBS. The pellets were washed twice with 1 ml of XB (100 mM KCl, 50 mM sucrose, 5 mM EGTA, 2 mM MgCl₂, 0.1 mM CaCl₂, 10 mM HEPES, pH 7.7) and incubated with 30 l of extracts for 1 h at 4°C. The supernatant was removed by centrifugation and treated as the immunodepleted extract. Aliquots of the pellets and supernatant were analyzed by immunoblotting. Purified G-actin from rabbit skeletal muscle was covalently labeled with tetramethylrhodamine iodoacetamide (Molecular Probes) as an actin tail tracer. Escherichia coli MC1060 ompT::Km carrying pD10-1 (Suzuki et al., 1996) was labeled with 4',6-diamidino-2-phenylindole (DAPI) (Sigma) and fixed with 4% paraformaldehyde in PBS. Cells were washed with XB and resuspended in buffer containing 20% glycerol. Bacterial motility was assayed by mixing 6 l of depleted extracts with 0.5 l each of DAPI-labeled bacteria and 0.5 mg/ml of rhodamine-labeled G–actin. Recombinant baculovirus expressing rat N-WASP was prepared by using the BAC-TO-BAC expression system (Gibco) (Miki et al., 1998). Cell lysates of infected Sf9 cells were applied to a HiTrap Heparin column (Pharmacia). After washing, the column was eluted with a gradient of 0–1000 mM NaCl. Purified protein was dialyzed against XB and concentrated by ultrafiltration.

Acknowledgements

We are grateful to Brigitte M.Jockusch and Jürgen Wehland for valuable discussions, and to Trinad Chakraborty for providing the L.monocytogenes strain. We thank Timothy J.Mitchison for helpful advice, Eisaku Katayama for image processing and Ruairí A.Mac Síomóin for critical reading of the manuscript. This work was supported by 'Research for the Future' Program of the Japan Society for the Promotion of Science, and Grant-in-aid for Scientific Research from the Japanese Ministry of Education, Science, Sports and Culture.

References

Allaoui A, Mounier J, Prévost C, Sansonetti PJ and Parsot C (1992) icsB: a Shigella flexneri virulence gene necessary for the lysis of protrusions during intercellular spread. Mol Microbiol, 6, 1605–1616. | PubMed | ISI | ChemPort |

Bernardini ML, Mounier J, D'Hauteville H, Coquis-Rondon M and Sansonetti PJ (1989) Identification of icsA, a plasmid locus of Shigella flexneri that governs bacterial intra and intercellular spread through interaction with F-actin. Proc Natl Acad Sci USA, 86, 3867–3871. | PubMed | ChemPort |

Brindle NPJ, Holt MR, Davies JE, Price CJ and Critchley DR (1996) The focal-adhesion vasodilator-stimulated phosphoprotein (VASP) binds to the proline-rich domain in vinculin. Biochem J, 318, 753–757. | PubMed | ChemPort |

Chakraborty T et al. (1995) A focal adhension factor directly linking intracellularly motile Listeria monocytogenes and Listeria ivanovii to the actin-based cytoskeleton of mammalian cells. EMBO J, 14, 1314–1321. | PubMed | ISI | ChemPort |

Derry JMJ, Ochs HD and Francke U (1994) Isolation of a novel gene mutated in Wiskott–Aldrich syndrome. Cell, 78, 635–644. | PubMed | ISI | ChemPort |

Domann E, Wehland J, Rohde M, Pistor S, Hartl M, Goebel W, Leimeister-Wachter M, Wuenscher M and Chakraborty T (1992) A novel bacterial gene in Listeria monocytogenes required for host cell microfilament interaction with homology to the proline-rich region of vinculin. EMBO J, 11, 1981–1990. | PubMed | ISI | ChemPort |

Gertler FB, Niebuhr K, Reinhard M, Wehland J and Soriano P (1996) Mena, a relative of VASP and Drosophila Enabled, is implicated in the control of microfilament dynamics. Cell, 87, 227–239. | PubMed | ISI | ChemPort |

Goldberg MB (1997) Shigella actin-based motility in the absence of vinculin. Cell Motil Cytoskel, 37, 44–53. | Article | ISI | ChemPort |

Goldberg MB and Theriot JA (1995) Shigella flexneri surface protein IcsA is sufficient to direct actin-based motility. Proc Natl Acad Sci USA, 92, 6572–6576. | PubMed | ChemPort |

Goldberg MB, Barzu O, Parsot C and Sansonetti PJ (1993) Unipolar localization and ATPase activity of IcsA, a Shigella flexneri protein involved in intracellular movement. J Bacteriol, 175, 2189–2196. | PubMed | ISI | ChemPort |

Haffner C, Jarchau T, Reinhard M, Hoppe J, Lohmann SM and Walter U (1995) Molecular cloning, structural analysis and functional expression of the proline-rich focal adhesion and microfilament-associated protein VASP. EMBO J, 14, 19–27. | PubMed | ChemPort |

Higley S and Way M (1997) Actin and cell pathogenesis. Curr Opin Cell Biol, 9, 62–69. | Article | PubMed | ChemPort |

Jockusch BM and Rüdiger M (1996) Crosstalk between cell adhesion molecules: vinculin as a paradigm for regulation by conformation. Trends Cell Biol, 6, 311–315. | Article | ChemPort |

Johnson RP and Craig SW (1994) An intramolecular association between the head and tail domains of vinculin modulates talin binding. J Biol Chem, 269, 12611–12619. | PubMed | ChemPort |

Johnson RP and Craig SW (1995) F-actin binding site masked by the intramolecular association of vinculin head and tail domains. Nature, 373, 261–264. | Article | PubMed | ISI | ChemPort |

Kadurugamuwa JL, Rohde M, Wehland J and Timmis KN (1991) Intercellular spread of Shigella flexneri through a monolayer mediated by membranous protrusions and associated with reorganization of the cytoskeletal protein vinculin. Infect Immun, 59, 3463–3471. | PubMed | ChemPort |

Kocks C, Gouin E, Tabouret M, Berche P, Ohayon H and Cossart P (1992) Listeria monocytogenes-induced actin assembly requires the actA gene product, a surface protein. Cell, 68, 521–531. | PubMed | ISI | ChemPort |

Kocks C, Marchand J-B, Goulin E, d'Hauteville H, Sansonetti PJ, Carlier M-F and Cossart P (1995) The unrelated surface proteins ActA of Listeria monocytogenes and IcsA of Shigella flexneri are sufficient to confer actin-based motility on Listeria innocua and Escherichia coli respectively. Mol Microbiol, 18, 413–423. | PubMed | ISI | ChemPort |

Laine RO, Zeile W, Kang F, Purich DL and Southwick FS (1997) Vinculin proteolysis unmasks an ActA homolog for actin-based Shigella motility. J Cell Biol, 138, 1255–1264. | Article | PubMed | ChemPort |

Lasa I and Cossart P (1996) Actin-based motility: towards a definition of the minimal requirements. Trends Cell Biol, 6, 109–114. | Article | ChemPort |

Lett M-C, Sasakawa C, Okada N, Sakai T, Makino S, Yamada M, Komatsu K and Yoshikawa M (1989) virG, a plasmid-coded virulence gene of Shigella flexneri: indentification of the virG protein and determination of the complete coding sequence. J Bacteriol, 171, 353–359. | PubMed | ChemPort |

Lowenstein EJ, Daly RJ, Batzer AG, Margolis B, Lammers R, Ullrich A and Schlessinger J (1992) The SH2 and SH3 domain-containing protein Grb2 links tyrosine kinases to ras signaling. Cell, 70, 431–442. | PubMed | ISI | ChemPort |

Machesky LM (1997) Complex dynamics at the leading edge. Curr Biol, 7, R164–R167. | PubMed | ChemPort |

Makino S, Sasakawa C, Kamata T and Yoshikawa M (1986) A genetic determinant required for continuous reinfection of adjacent cells on a large plasmid in Shigella flexneri 2a. Cell, 46, 551–555. | PubMed | ISI | ChemPort |

Matuoka K, Shibata M, Yamakawa A and Takenawa T (1992) Cloning of ASH, a ubiquitous protein composed of one Src homology region (SH) 2 and two SH3 domains, from human and rat cDNA libraries. Proc Natl Acad Sci USA, 89, 9015–9019. | PubMed | ChemPort |

Miki H, Miura K, Matuoka K, Nakata T, Hirokawa N, Orita S, Kaibuchi K, Takai Y and Takenawa T (1994) Association of Ash/Grb-2 with dynamin through the Src homology 3 domain. J Biol Chem, 269, 5489–5492. | PubMed | ISI | ChemPort |

Miki H, Miura K and Takenawa T (1996) N-WASP, a novel actin-depolymerizing protein, regulates the cortical cytoskeletal rearrangement in a PIP2-dependent manner downstream of tyrosine kinases. EMBO J, 15, 5326–5335. | PubMed | ISI |

Miki H, Sasaki T, Takai Y and Takenawa T (1998) Induction of filopodium formation by a WASP-related actin-depolymerizing protein N-WASP. Nature, 391, 93–96. | Article | PubMed | ISI | ChemPort |

Nakata N, Tobe T, Fukuda I, Suzuki T, Komatsu K, Yoshikawa M and Sasakawa C (1993) The absence of surface protease, OmpT, determines the intercellular spreading ability of Shigella: the relationship between the ompT and kcpA loci. Mol Microbiol, 9, 459–468. | PubMed | ISI | ChemPort |

Niebuhr K et al. (1997) A novel proline-rich motif present in ActA of Listeria monocytogenes and cytoskeletal proteins is the ligand for the EVH1 domain, a protein module present in the Ena/VASP family. EMBO J, 16, 5433–5444. | Article | PubMed | ISI | ChemPort |

Nobes CD and Hall A (1995) Rho, rac, and cdc42 GTPases regulate the assembly of multimolecular focal complexes associated with actin stress fibers, lamellipodia, and filopodia. Cell, 81, 53–62. | PubMed | ISI | ChemPort |

Pantaloni D and Carlier M-F (1993) How profilin promotes actin filament assembly in the presence of thymosin beta 4. Cell, 75, 1007–1014. | PubMed | ISI | ChemPort |

Prévost M-C, Lesourd M, Arpin F, Vernel F, Mounier J, Hellio R and Sansonetti PJ (1992) Unipolar reorganization of F-actin layer at bacterial division and bundling of actin filaments by plastin correlate with movement of Shigella flexneri within HeLa cells. Infect Immun, 60, 4088–4099. | PubMed  |

Rajakumar K, Jost BH, Sasakawa C, Okada N, Yoshikawa M and Adler B (1994) Nucleotide sequence of the rhamnose biosynthetic operon of Shigella flexneri 2a and role of lipopolysaccharide in virulence. J Bacteriol, 176, 2362–2373. | PubMed | ChemPort |

Reinhard M, Giehl K, Abel K, Haffner C, Jarchau T, Hoppe V, Jockusch BM and Walter U (1995) The proline-rich focal adhesion and microfilament protein VASP is a ligand for profilins. EMBO J, 14, 1583–1589. | PubMed | ISI | ChemPort |

Reinhard M, Rüdiger M, Jockusch BM and Walter U (1996) VASP interaction with vinculin: a recurring theme of interactions with proline-rich motifs. FEBS Lett, 399, 103–107. | Article | PubMed | ChemPort |

Rosenblatt J, Agnew BJ, Abe H, Bamburg JR and Mitchison TJ (1997) Xenopus actin depolymerizing factor/cofilin (XAC) is responsible for the turnover of actin filaments in Listeria monocytogenes tails. J Cell Biol, 136, 1323–1332. | Article | PubMed | ISI | ChemPort |

Sandlin RC, Lampel KA, Keasler SP, Goldberg MB, Stolzer AL and Maurelli AT (1995) Avirulence of rough mutants of Shigella flexneri: requirement of O antigen for correct unipolar localization of IcsA in the bacterial outer membrane. Infect Immun, 63, 229–237. | PubMed | ChemPort |

Sanger JM, Sanger JW and Southwick FS (1992) Host cell actin assembly is necessary and likely to provide the propulsive force for intracellular movement of Listeria monocytogenes. Infect Immun, 60, 3609–3619. | PubMed | ISI | ChemPort |

Sasakawa C, Kamata K, Sakai T, Murayama Y, Makino S and Yoshikawa M (1986) Molecular alteration of the 140-megadalton plasmid associated with loss of virulence and congo red binding activity in Shigella flexneri. Infect Immun, 51, 470–475. | PubMed | ISI | ChemPort |

Suzuki T, Murai T, Fukuda I, Tobe T, Yoshikawa M and Sasakawa C (1994) Identification and characterization of a chromosomal virulence gene, vacJ, required for intercellular spreading of Shigella flexneri. Mol Microbiol, 11, 31–41. | PubMed | ChemPort |

Suzuki T, Lett M-C and Sasakawa C (1995) Extracellular transport of VirG protein in Shigella. J Biol Chem, 270, 30874–30880. | Article | PubMed | ISI | ChemPort |

Suzuki T, Saga S and Sasakawa C (1996) Functional analysis of Shigella VirG domains essential for interaction with vinculin and actin-based motility. J Biol Chem, 271, 21878–21885. | Article | PubMed | ChemPort |

Theriot JA, Mitchison TJ, Tilney LG and Portnoy DA (1992) The rate of actin-based motility of intracellular Listeria monocytogenes equals the rate of actin polymerization. Nature, 357, 257–260. | Article | PubMed | ISI | ChemPort |

Theriot JA, Rosenblatt J, Portnoy DA, Goldschmidt-Clermont PJ and Mitchison TJ (1994) Involvement of profilin in the actin-based motility of L.monocytogenes in cells and in cell-free extracts. Cell, 76, 505–517. | PubMed | ISI | ChemPort |

Welch MD, Mallavarapu A, Rosenblatt J and Mitchison TJ (1997) Actin dynamics in vivo. Curr Opin Cell Biol, 9, 54–61. | Article | PubMed | ISI | ChemPort |

Zeile WL, Purich DL and Southwich FS (1996) Recognition of two classes of oligoproline sequences in profilin-mediated acceleration of actin-based Shigella motility. J Cell Biol, 133, 49–59. | PubMed | ISI | ChemPort |

Zigmond SH (1996) Signal transduction and actin filament organization. Curr Opin Cell Biol, 8, 66–73. | Article | PubMed | ISI | ChemPort |